I have attached two PDFs (links below) that discuss the effect of Marangoni Negative systems.  It is generally agreed that HETP is at least two times greater for Marangoni negative systems but I am not sure that that is the whole story as you will see from the Koshy- Rukovena paper and the effect that changes in lambda has on the HETP.  A high lambda system with a Marangoni negative affect is going to have an HETP increased at least in the range indicated by Kister.  I am sorry but I don’t know how to predict the HETP closer than that.  Note from Figure 3 of the Koshy-Rukovena paper that a system with a high or low lambda can have its HETP increase be even greater than three times.

http://www.chemicalprocessing.com/Media/MediaManager/Distillation_columns_Zuiderweg_Harmens002.pdf

http://www.chemicalprocessing.com/Media/MediaManager/Reflux_and_Surface_Tension_Effects_Koshy_and_Ruko001.pdf